1 month ago
San Jose, CA, USAMid Level
Base Salary
$200k - $400k/yr
Responsibilities
- Design, train, evaluate, and deploy learning-based visuomotor manipulation policies for humanoid robots.
- Develop manipulation behaviors including grasping, pick-and-place, object reorientation, door opening, bimanual manipulation, and basic assembly.
- Apply and extend behavior cloning, reinforcement learning, and VLA reasoning techniques.
- Train robust models that handle sensor noise, partial observability, contact dynamics, and environmental variability.
- Own the pipeline from real-robot data collection through model training, evaluation, and deployment.
- Collaborate with simulation, perception, controls, systems, hardware, integration, and testing teams.
- Evaluate learning-based and classical robotics approaches and make principled design decisions.
- Write well-tested software that runs reliably on physical humanoid robots.
- Improve system robustness, performance, and deployment velocity.
Requirements
- Hands-on experience developing and deploying robot learning systems on real robots.
- Strong background in robot manipulation and visuomotor control.
- Experience with behavior cloning, reinforcement learning, or related learning-based manipulation methods.
- Proficiency in Python and/or C++ for robotics and machine learning systems.
- Experience with modern deep learning frameworks such as PyTorch.
- Ability to design experiments, analyze failures, and iterate on real-world robotic systems.
- Understanding of tradeoffs between classical robotics and learning-based methods.
- Bonus: experience deploying manipulation systems in commercial or production robotics, humanoid or highly dexterous robotics experience, publications in robot learning or embodied AI, and experience leading projects or mentoring engineers.
Benefits
- Full-time position.
Categories
ML EngineeringRobotics